This project will assemble, collect and analyze multiple rounds of survey data from Bangladesh, providing family-based and individual panel information on the long-term health and productivity effects of childhood nutritional intakes, indoor air pollution, and health interventions over a 25-year span, using a newly-available panel survey as well as a proposed additional survey round. The completed panel survey is well-suited for investigating the long-term relationships between childhood health, other human capital investments, important dimensions of adult productivity, adult health, migration, and marriage-market outcomes because the survey design minimizes the biases due to two major sources of selectivity in many existing surveys: selectivity of individuals with respect to spatial mobility and selectivity of outcome measures by activity choice. The panel data set has been designed to enable the best estimates possible from non-experimental data of the effects of early nutritional and other health interventions on adult outcomes when (i) resource allocations to children are choices made by optimizing parents within a family context, (ii) there is unobserved heterogeneity in health endowments, and (iii) adults are geographically mobile. The methods of analysis we will use will exploit the combination of three important dimensions of the data: (i) information from multiple time periods for the same individuals, (ii) information on siblings (of any age), and (iii) information on the locations of the respondents. The analyses will also reveal aspects of the interrelationships between pollution exposure and nutrition and their effects on child and adult health that are absent in most of the literature due to inattention to health heterogeneity and inadequate data.